09-18-07 - 7

The main advantage of privatization is that it wipes the slate of regulation. It's not because the private enterprise is "more efficient" - a privatized public operation works in a subsidized cocoon free of market pressure, it's not part of the magic capitalist efficiency machine. Rather, it gets to severely cut quality, and thus cut costs. The sneaky thing is this quality cut is not always obvious, because it's often a cut in disaster preparedness or safety or environmental cleanliness, etc. Over the years of a public institutions' existance, the regulations build up as different people spot problems and pass laws to handle them. Certainly you often wind up with too much regulation, but most of the time that is still better than none at all. Politically, you can't strip away those regulations because people can easily see them and object, "we need the safety, we need the quality", so just privatizing the whole thing sounds great and people don't realize that they are all losing all the quality and operations protections.

In general, preventing rare problems is incredibly inefficient and private enterprise doesn't want to do it. One of the interesting things to me about the Minnesota bridge collapse is that there was this little subtext that nobody wanted to say publicly : our infrastructure maintenance program is specifically designed so that we have an occasional bridge collapse. That is, we have a very limited budget for new bridges and repairs, so things are evaluated for safety and the ones that are obviously critical are repaired, but it's all about percentages. One bridge might be evaluated as a 0.1% chance of failing next year, then they have to decide whether that's worth repairing. In private enterprise you would evaluate the cost of preventing the accident to the cost if it happens, in terms of all the suits you'd have to pay and the negative publicity. Usually what you find is a sort of exponential decay of failure rates to constant expenditure. That is, say you're running an oil refinery. There's a pretty high chance of some kind of industrial accident injuring someone. Maybe you estimate the current chance is around 1% and the cost to halve that chance is about $1 Million. Okay, you pay that. Now the chance is 0.5% and the chance to have that is $1 M again. Okay, go ahead and pay that. But pretty soon you get to a point of diminishing returns and you just want to accept the small chance of failure.

Of course those higher standards can also really screw up public enterprise. Say you're running a bussing company on a limited budget. The private company might decide to just run a fleet of busses along the main commuter routes, stopping every 5 blocks only on major roads. Not everyone can use them, but they get a lot of traffic and the users are quite happy. The public company is forced by voters to run busses all over the city, stopping every 2 blocks; they're now accessible to all but everyone thinks the service sucks.

There are of course also other tricks that the government can use to wipe regulation. One is making a brand new institution or agency, then having it gradually usurp responsibilities from old agencies, but under new names.

No comments:

old rants